www.nature.com/scientificreports

OPEN Estimating the ecological water levels of shallow lakes: a case study in Tangxun Lake, Wei Yang1,2*, Mingxiang Xu1, Ruiqing Li1, Liping Zhang2* & Qiuliang Deng1

Water level management is an efective tool for the ecological restoration of shallow lakes. In this study, we developed an ecologically-based approach to estimate the monthly suitable ecological water levels (EWLs). This approach took both the lake topographic features and aquatic plants’ growth characteristics into account. The aquatic vegetation coverage was used to characterize the degree of the lake ecological restoration. The relationship between water level and vegetation coverage was established. We chose the Tangxun Lake as a testbed, and the recommended lowest EWL was 16.6 m, as the minimum threshold for water level regulations. The results revealed that the predicted vegetation coverage decreased with the rise of water level during the germination period (February and March). To achieve the vegetation coverage goal of 30% and 50%, the lake’s water levels must be lowered to 17.1 m and 16.8 m respectively during germination. The EWLs were recommended to be low in spring and high in summer, which was matched with the natural water level regimes. The proposed approach can provide a reliable reference for water level regulation of shallow lakes especially the lakes with insufcient data.

Lake is an important part of the natural ecosystem and plays a vital role in human’s survival and development. In the past decades, the hydrological regime of lakes has changed signifcantly in China due to the synergies of human activities and climate change, leading to a series of ecological problems, such as water environment dep- ravation, eutrophication aggravation and lakeside habitat destruction1–3. To address these problems, the govern- ment has tried numerous measures including strict control of point source pollution and some bioremediation technologies. Additionally, water level manipulation can be considered an efective tool for wetland restoration4,5. Water level is an important characteristic index to refect the lakes’ hydrological regime, as well as a key factor infuencing the distribution and diversity of aquatic plants in shallow lakes6,7. To maintain the lakes’ basic ecolog- ical functions, the concept of lake ecological water level (EWL) was developed. And it was defned as the optimal water level for maintaining the ecosystem integrity, protecting the biodiversity, improving the environmental quality and ensuring the ecosystem stability8,9. A thorough understanding of EWL is critical to restoring the lake ecosystems10. Tere has been an increasing number of studies on the EWL of lakes, and these studies mainly focus on two aspects: (1) the infuences of water level fuctuations on the establishment and development of biological com- munities in lakes; (2) the active or passive responses of biological communities to the water level fuctuations of lakes. Water level fuctuations, especially their extent, frequency and duration, have great efects on the lake’s physical environment and biological communities6,11. Particularly, since aquatic plants play a vital role in the maintenance of lake ecosystem health, much research has been undertaken to investigate the impacts of water level fuctuations on aquatic plants by using ecological indicators such as biodiversity and coverage12–14. It has been found that minor changes in water levels could produce huge alterations in plant communities15. Moreover, the rates of water level change are also dominant forces afecting the growth and development of aquatic plants. Aquatic plants gradually adapt to the natural hydrological regime in the long process of evolution, and they have diferent water level requirements at diferent growth stages16. Based on this, many countries have attempted to formulate water level manipulation schemes to restore natural water level regimes and ensure the species diversity and ecosystem stability of lake wetlands.

1Hubei Provincial Water Resources and Hydropower Planning Survey and Design Institute, , 430064, China. 2State Key Laboratory of Water Resources and Hydropower Engineering Science, , Wuhan, 430072, China. *email: [email protected]; [email protected]

Scientific Reports | (2020)10:5637 | https://doi.org/10.1038/s41598-020-62454-5 1 www.nature.com/scientificreports/ www.nature.com/scientificreports

How to determine the appropriate EWLs in shallow lakes has become a new research hotspot. Since the end of the 20th century, a variety of methods have been developed for calculating the EWLs. Te scope of the EWLs assessment methods has gradually broadened from the hydrological analyses, towards more comprehen- sive approaches. Commonly used methodologies fall into four categories: historical fow record methodolo- gies, hydraulic rating methodologies, habitat rating methodologies, and holistic methodologies17. Te historical fow record methodologies and the hydraulic rating methodologies rely heavily on the availability of reliable long-term hydrological data9,18. Tese two methodologies are based on the concept that water levels below the natural water levels will destroy the ecosystem integrity19. Te habitat rating methodologies were derived from the hydraulic rating methodologies and were widely used in recent years. Tey draw on the ecological require- ments of indicator species to assess the suitability for ecosystem processes. And the common approaches include the Physical Habitat Simulation Model20, the Minimum Biological Space Requirements Method21, the Instream Flow Incremental Methodology22 and the Habitat Analysis Method23. Te holistic methodologies are comprehen- sive multidisciplinary approaches considering both hydrological and ecological indicators. Since the interaction mechanism between hydrological factors and ecological factors is extremely complicated, the research of this methodology is still at its initial period24. Among the existing methodologies, most of them were developed to determine the environmental fow requirements of rivers, while they could not directly applicable to lake wet- lands. Moreover, the primary focuses of these methods have been on the lowest EWLs, with less emphasis on the efects of dynamic water level fuctuations on the aquatic species. However, a single EWL value could not meet the water level requirements at diferent growth stages of aquatic species25,26. Te extent, frequency and duration of water levels are more worthy of attention. Terefore, further studies are still necessary to develop the methods for EWLs calculation in lakes that could meet the water level requirements at diferent growth stages of aquatic species. In this study, we developed an approach for calculating the suitable monthly EWLs of shallow lakes. Te essence of this approach was to regulate the water level based on the water level requirements of aquatic plants. Te vegetation coverage was used as an indicator of ecological health in this approach. Te lowest EWL was cal- culated as the minimum threshold for water level regulation using three approaches. Te principle and procedure of this method were introduced in detail, and then the methodology was applied to the Tangxun Lake. Materials and Methods Study area. In this study, we selected the Tangxun Lake in China as the study area. It is the largest urban lake in China and located in the middle and lower reaches of the River (Fig. 1). Te Tangxun Lake has a water surface area of 52.19 km2 and drains a watershed area of 240.38 km2. It is a typical subtropics shallow lake with an average water depth of 1.85 m. It serves as the main water source of drinking, irrigation and aquaculture for Wuhan, the capital city of Province. Te Tangxun Lake used to be the largest original ecological lake in Wuhan. However, in recent years, the increase of population and economic development in the basin, especially the construction of industrial parks and development zones around the basin since 1996, has discharged a large amount of pollutant load that exceeds the environmental capacity of the water body, resulting in water pollution and eutrophication in the lake. As a result, the biological diversity and the aquatic vegetation coverage are sharply reduced, which poses a serious threat to the health of the surrounding residents.

Traditional methods for estimating EWLs. Te lowest EWLs are generally emphasized in traditional methods of estimating the EWLs, and are usually estimated through the following approaches.

Lake morphological analysis method (LMAM). Te LMAM was proposed by Xu et al. and was widely used in China27. In this method, water level is used as the index of lake topography and hydrological condition, and lake area is used as the index of lake function. Based on the measured water level and lake area data, the relation curve between water level and lake area can be established. Te change rate of lake area is the frst derivative of the rela- tion function between lake area and water level. Te water level corresponding to the maximum change rate of the lake area is treated as the lowest EWL of the lake. A major assumption of this method is that if the water level is lower than the lowest ecological water level, the surface area of the lake will be signifcantly reduced and the lake function will be seriously degraded. Tis method can be expressed as:

Ff= ()H (1)

∂2F = 0 ∂H2 (2) where F is lake area (m2); H is water level (m). Te lowest EWL can be obtained by solving the equations.

Natural water level statistics (NWLS). Some researchers demonstrated that the annual and inter-annual changes in water level cause disturbance to the lake ecosystem under natural conditions21. Te premise of the NWLS is that the lake ecosystem has adapted to the disturbance of lake level during the long ecological evolution28. Te long-term daily water level data is required in this method, then the water level guarantee rate curve can be plot- ted. Te water level with guarantee rate of 95% is generally considered as the lowest EWL29,30.

Biological living space requirement method (BLSRM). Te aquatic organisms in lakes include phytoplankton, emergent plants, zooplankton, fshes and so on. Each biological community needs a minimum living space to

Scientific Reports | (2020)10:5637 | https://doi.org/10.1038/s41598-020-62454-5 2 www.nature.com/scientificreports/ www.nature.com/scientificreports

Figure 1. Location sketch of study area and its underwater terrain map.

maintain its own community from severe recession. Te water level corresponding to this living space is the low- est EWL of the lake. Te key issue is to identify the organism most sensitive to the water level and then determine the lowest water level that the organism requires to survive and reproduce. Te lowest EWL can be calculated as:

HHmin =+bch (3)

where Hmin is the lowest EWL of the lake; Hb is the bottom elevation of the lake; hc is the lowest water depth that the organism requires.

The proposed approach for estimating the suitable EWLs. Fundamental principles of the approach. Te aquatic vegetation coverage, which is defned as the percentage of aquatic vegetation area in the total area of the lake, is a very important index to score the growth conditions of aquatic plants in the lake ecosystems. In this approach, it was used as the main ecological restoration target in the regulation of ecological water level. Te growth and reproduction of aquatic vegetation are closely related to the water level fuctuation. Te growth periods of aquatic plants in subtropical lakes are generally divided into six stages: germination (February-March), seedling growth (April-May), growth and difusion (June-July), maturation (August-September), seed propaga- tion (October-November) and dormancy (December-January). Among these growth periods, the germination stage is particularly important since it determines the plant distribution and vegetation coverage of the lakes. Terefore, February and March is regarded as the critical period of aquatic vegetation restoration and the bench- mark period of water level regulation. Ten the suitable EWLs of the lake can be obtained based on the water depth and water level variation requirements of indicator species.

Calculation method framework. Te framework of this approach was shown in Fig. 2. Te lowest EWL was considered as the minimum threshold level for maintaining the functional integrity and biodiversity of the lake ecosystems, and it was calculated using the methods mentioned above. Aferwards, the relation curve between water level and its corresponding water area in lakes was plotted by linking water level data with underwater topography data. Furthermore, a specifc quantitative relationship between vegetation coverage and water levels during germination was established based on the water requirements for aquatic plants. Based on this, the recom- mended water levels during germination were derived once the vegetation coverage objectives were determined. Ten the suitable EWLs of other months could be estimated according to the requirements of aquatic plants in water-level changing speed and water depth.

Scientific Reports | (2020)10:5637 | https://doi.org/10.1038/s41598-020-62454-5 3 www.nature.com/scientificreports/ www.nature.com/scientificreports

Figure 2. Te framework of the proposed approach.

Months Growth stages Water level requirements February-March Germination Low water level, higher than the lowest EWL Gradually increased water level, the rising speed April-May Seedling growth must be less than 0.6 m/month High water level, the rising speed must be less than June-July Growth and Difusion 0.7 m/month High water level, lower than the warning water August-September Maturation level Gradually decreased water level, the drawdown October-November Seed propagation speed must be less than 3 cm/day December-January Dormancy Low water level

Table 1. Water level requirements of aquatic vegetation during each growth period.

Water level requirements of aquatic plants. Aquatic plants in lake ecosystems mainly include several types, such as emergent plants, hygrophytes, foating-leaved plants and submerged plants, and most of them are located in the lakeside zone. Te emergent plants and hygrophytes can both germinate in shallow areas of the lake, while the foating-leaved plants and submerged plants are usually found under water. Previous research showed that submerged plants could develop only when the ratio of Secchi Depth (SD) to water depth was beyond 0.6, and the emergent plants could develop where the water depth was less than 20 cm16. Most of the aquatic vegetation in Tangxun Lake germinates in February and March. During this time, the lake need to keep a low water level to increase the exposed beach area, and the specifc calculation process was shown in the following section. Te Phragmites communis community is the predominant community and have wide distributions in the Tangxun Lake wetlands. Te growth rate of the Phragmites communis was about 0.7 m/month, and the mean height was 0.6 m in April, 1.0 m in May, 2.2 m in June, and 2.8 m in July, respectively, and stopped growing afer August31,32. To ensure the normal growth and development of the aquatic plants, the water level must not exceed the tops of the Phragmites communis. In the seedling growth period (April and May), it is necessary to keep the water level rising steadily and slowly, and the rate of increase should be controlled within 0.6 m/month (Table 1). In the growth and difusion period (June and July), the lake is suitable to maintain a high water level, but the rising speed should not exceed 0.7 m/ month. Maintaining a high water level can not only promote the spread of aquatic plants, but also prevent the lakeside from shrinking. To prevent terrestrial plant invasion and lake swamping, the lake water level is better to keep a high value during the maturation period (August and September), but it should not exceed the warn- ing water level of the lake. During the seed propagation period (October and November), the lake is suitable to maintain a moderate water level. To promote the maturation and spread of seeds, it is necessary to keep the water level steady and slowly decline, and the rate of decline should not exceed 3 cm/day. During the dormancy period (December and January), the water level should maintain a medium or low value.

Calculation steps. Te distribution of aquatic plants in a wetland is primarily a function of water depth14,33, and the coverage can be calculated according to the water level during the germination period, and the calculation steps are as follows: Step 1: Establish a relationship between water level (Z) and lake surface area (A). A series of water surface areas (A) corresponding to water levels (Z) can be derived from the underwater ter- rain data of the lake, and then the function relation between Z and A can be obtained and expressed as A = f(Z). Step 2: Calculate the aquatic vegetation coverage of the lake.

Scientific Reports | (2020)10:5637 | https://doi.org/10.1038/s41598-020-62454-5 4 www.nature.com/scientificreports/ www.nature.com/scientificreports

Te exposed beach area from February to March was used as the germination and growth area of hygrophytes and emergent plants, and that can be calculated as the surface area between the normal water level (Zc) and the water level during germination (Zg). In addition, previous research showed that the emergent plants could also develop where the water depth was less than 20 cm. Terefore, the elevation distribution of hygrophytes and emergent species was from (Zg − 0.2 m) to Zc. It was found that the submerged plants could develop only when the ratio of Secchi Depth (SD) to water depth was beyond 0.6, so the lowest elevation of submerged plants was calculated as Zg − SD/0.6. Te elevation distribution of submerged plants was from (Zg − SD/0.6) to Zg, which was partially overlapped with hygrophytes and emergent plants. Therefore, the elevation distribution of submerged plants, hygrophytes and emergent species was (Zg − SD/0.6)~Zg, Zg ~ Zc and (Zg − 0.2 m) ~ Zc, respectively. Te lowest elevation of aquatic plants was the minimum value of Zg − SD/0.6 and Zg − 0.2 m, and the elevation distribution of aquatic plants was from min((Zg − SD/0.6), (Zg − 0.2 m)) to Zc. Te corresponding planimetric area, that was the germination area of aquatic plants, can be calculated according to the function A = f(Z). Hence the coverage of the lake was calculated as:

AA−−min,ZASD Z −.02 cg06. g C = ((() )) × 100% Ac fZ()−−min,fZ SD fZ−.02 cg06. g = ((() )) × 100% fZ()c (4) Step 3: Establish a relationship between vegetation coverage (C) and water level during germination (Z). Repeating the above calculating steps, the coverage at any germination water level could be calculated. Zg was assigned a value from Zmin to Zc, in increments of 0.1 m, where Zmin was the lowest EWL of the lake, and it can be calculated according to the method mentioned above. Ten the relationship between vegetation coverage and water level (FC~Z) can be established. Step 4: Calculate the water level during germination under a given coverage target. To determine the suitable EWLs under target coverage, the water level at the germination stage should be calculated frst. Te restoration target of the aquatic vegetation coverage depends on the management goal of the lake. Once the aquatic vegetation coverage is determined, the water level requirement during germination can be calculated using the function FC~Z. Step 5: Determine the suitable EWLs for other growth stages. Tree objectives must be taken into account in the water level manipulation. Te frst was to rehabilitate the aquatic vegetation coverage. A lowered water level may provide more suitable conditions for germination and seed- ling growth of hygrophytes and emergent species. Actually, previous studies indicated that the trends of increasing coverage corresponded with low water levels, and decreasing coverage corresponded with high water level34. Hence, water level must keep low in spring, but not lower than the lowest EWL. Te second was to ensure the normal growth and development of aquatic plants. To achieve this objective, water level requirements of aquatic vegetation in each growth periods should be taken into account. Te third was to guarantee the safety of the lakes during food season. In summer, high water levels were favourable for the dispersal of aquatic plant seeds, and they could also limit exotic vegetation encroachment. However, to guarantee the safety of the lakes, water level should not higher than the warn- ing water level. Based on the above objectives, the upper and lower limits of the water levels can be obtained. Results Calculation of the lowest EWLs. Te lowest EWL calculated by LMAM. According to the underwater topography data of the Tangxun Lake, the relationship curve between water level (Z) and lake surface area (A) was plotted, as shown in Fig. 3(a). Ten a series of the change rate of the lake surface area (dF/dZ) corresponding to water levels (Z) was derived, and the relation graph was displayed in Fig. 3(b). In this graph, the change rate of the lake surface area reached the maximum value when the water level was 15.95 m, so the lake’s lowest EWL calculated by LMAM was 15.95 m.

Te lowest EWL calculated by NWLS. Tis paper selected the measured water levels of the lake from 1952 to 1987, which were less afected by artifcial regulation and approximate to the natural water level process. Te water levels under diferent guarantee rates can be obtained by arranging the water level data from 1952 to 1987 in the order from small to large, as shown in Fig. 4. It can be discerned from the chart that the water level with guarantee rate of 95% was 16.48 m, so the lake’s lowest EWL calculated by NWLS was 16.48 m.

Te lowest EWL calculated by BLSRM. Fish has a unique position in the aquatic ecosystem and is one of the top communities in the aquatic ecosystem, which plays an important role in the existence and abundance of other groups. Compared with other biological groups, fsh is more sensitive to low water levels, thus fsh is selected as the indicator species. Previous studies have indicated that the minimum water depths required for lacustrine fshes to survive are about 1.0 m. According to the underwater topography data, the bottom elevation of the Tangxun Lake was 15.6 m. Terefore, the lowest EWL calculated by BLSRM was 15.6 + 1.0 = 16.6 m. By using LMAM, NWLS and BLSRM, the lowest EWL was calculated as 15.95 m, 16.48 m and 16.6 m respec- tively. For safety considerations, the lowest EWL of the Tangxun Lake was recommended as 16.6 m.

Calculation of the suitable EWLs. Before 2000, the aquatic vegetation coverage of the Tangxun Lake was more than 30%, while afer 2013, it was decreased to 10% due to the degradation of the aquatic plants especially the submerged plants. In this paper, the target vegetation coverage was set to be 30~50%.

Scientific Reports | (2020)10:5637 | https://doi.org/10.1038/s41598-020-62454-5 5 www.nature.com/scientificreports/ www.nature.com/scientificreports

60 45 ) (a) 40 (b) /m 2

) 50 2 m 35 m (k (k 40 30 area

area 25 30 la ke 20 of 20

surface 15

rate 10 la ke 10 5 0 change 0 15 15.5 16 16.5 17 17.5 18 18.5 19 15 15.5 16 16.5 17 17.5 18 18.5 19 water level (m) water level (m)

Figure 3. Relationships between water level and (a) lake area and (b) change rate of lake area.

20.5 20 19.5 ) 19 (m l

ve 18.5 le 18 er 17.5 wat 17 16.5 16 020406080100 guarantee rate (%)

Figure 4. Te water level guarantee rate curve. Te water level guarantee rate represents the probability that the water level is greater than a certain value.

According to the measured data, the mean Secchi depth (SD) in February and March was about 50 cm, and the normal water level (Zc) was 17.63 m. Te lowest EWL of the Tangxun Lake was 16.6 m based on the calculation results in the previous section, so frst we assigned the water level during germination period (Zg) to 16.6 m. In this case, the elevation distribution of submerged plants, hygrophytes and emergent species was 15.8~16.6 m, 16.6~17.63 m and 16.4~17.63 m, respectively. Ten the coverage (C) of the lake was calculated to be 70.5% by the Eq. (4), which means the coverage of the lake could reach to 70.5% when the water level during germination period was 16.6 m. Repeating the above calculating steps, Zg was assigned a value from 16.6 m to 17.63 m (16.6 m, 16.7 m, …, 17.7 m), in increments of 0.1 m, and the coverage at each germination water level can be calculated, as shown in Fig. 5. Te graph indicated that the vegetation coverage decreased with the germination water level rose. Hence a model to predict the coverage of the lake at any potential germination water level was developed by a polynomial ftting method, as expressed to: C = 34.313Z2 − 1230.1Z + 11035 (16.6 ≤ Z ≤ 17.7), where C was the coverage (%) and Z was the water level during germination period (m). Based on this model, the water levels required during the germination stage (February and March) were about 17.1 m and 16.8 m respectively under the coverage scenarios of 30% and 50%. And the predicted germination zones of aquatic plants in lakes were located in the area with low water depth (Fig. 6). Based on the water level requirements of aquatic vegetation (Table 1), the recommended suitable EWLs during each growth period can be obtained (Fig. 7). It can be seen from the chart that the water levels fuctuated between the lowest EWL (16.6 m) and the warning water-level (20.5 m). A similar variation trend of the suitable EWLs was observed under the coverage scenarios of 30% and 50%. To achieve the vegetation coverage of 30%, the lake’s water levels remained steady at 17.1 m during germination period (February and March). In the seedling growth period (April and May) and growth difusion period (June and July), the water level rose slowly, and reached the top in August. Tere was a gradual decline during the seed propagation period (October and November) and the dormancy period (December and January). Discussion and Conclusion The relationship between vegetation coverage and water level. Although water level management is an efective tool for the ecological restoration of shallow lakes, the relationship between ecosystem health and water level have rarely been quantifed. In this paper, we used the aquatic vegetation coverage to indicate the health of the ecosystem, and then established the quantitative relationship between vegetation coverage (C) and water level during germination (Z) in the Tangxun Lake, which was expressed as C = 34.313Z2 − 1230.1Z + 11035 (16.6 ≤ Z ≤ 17.7). Te results showed that vegetation coverage decreased as the germination water level increased. Te empirical relationship can be used to predict the potential aquatic vegetation coverage under a

Scientific Reports | (2020)10:5637 | https://doi.org/10.1038/s41598-020-62454-5 6 www.nature.com/scientificreports/ www.nature.com/scientificreports

80 2 70 y = 34.313x -1230.1x + 11035 60 ) 50 (% 40 30 coverage 20 10 0 16.5 16.7 16.9 17.1 17.3 17.5 17.7 water lever during germination (m)

Figure 5. Relationship between vegetation coverage and water level during germination in Tangxun Lake. Points represent the vegetation coverage at each germination water level. Line represents the ftted curves of the relation between vegetation coverage and germination water level.

Figure 6. Te predicted germination zones of aquatic plants in lakes with the water level of (a) 17.1 m and (b) 16.8 m during germination.

(a) upper limit lower limit (b) upper limit lower limit

21 warning water-level 21 warning water-level

20 20 ) ) l (m 19 l (m 19 ve ve le normal water-level le 18 18 normal water-level ter ter wa 17 wa 17 lowest EWL 16 lowest EWL 16

15 15

Figure 7. Te recommended monthly EWLs to achieve vegetation coverage of (a) 30% and (b) 50%. Te solid lines represent the upper limits of the EWLs, and the dotted lines represent the lower limits of the EWLs.

given water level condition and to help defne restoration targets. To maintain the vegetation coverage at 30%, the water level of the Tangxun Lake during germination period must be lowered to 17.1 m. To achieve the vegetation coverage goal of 50%, the water level during the germination period must be lowered to 16.8 m. Terefore, lowering water level during germination (February and March) is an efective measure for aquatic vegetation restoration. To further confrm this conclusion, we collected the water level and aquatic vegetation coverage of Eastern Taihu Lake from 2003 to 201735 and established a relationship between these two variables, as shown in the Fig. 8. It can be seen that aquatic vegetation coverage and water levels during germination in the Eastern Taihu Lake also presented a negative relationship. In addition, many previous experiment and feld stud- ies have shown a notable negative correlation between aquatic vegetation coverage and water level in spring35–37.

Scientific Reports | (2020)10:5637 | https://doi.org/10.1038/s41598-020-62454-5 7 www.nature.com/scientificreports/ www.nature.com/scientificreports

100 95 y = 282.31x2 -1874x + 3183.6 R² = 0.6969 90 )

(% 85 80 75 coverage 70 65 60 33.053.1 3.15 3.23.253.3 3.35 water level during germination (m)

Figure 8. Relationship between vegetation coverage and water level during germination in Eastern Taihu Lake.

For example, Wu36 conducted the experiments with submerged macrophytes in diferent water levels, and the results showed that high water level in spring was not conducive to the germination of submerged plants due to insufcient underwater light. Terefore, we strongly recommend that the water level of lakes be appropriately lowered in spring on the premise of ensuring the stable basic functions so as to facilitate the germination of aquatic plants.

Advantages of the proposed approach for estimating the suitable EWLs. It is increasingly rec- ognized that water level management is an efective tool for the ecological restoration of shallow lakes. To deter- mine the specifc and feasible management measures for the preservation and restoration of aquatic vegetation, a reliable and fexible method for estimating the EWLs need to be put forward. However, most of the traditional methods focus on the lowest EWL and ignore the efects of water-level fuctuation on aquatic organisms. In this paper, we proposed a simple methodology based on the water level requirement of aquatic plants, which helped stakeholders to determine the lake levels for each period of the year. Depending on the shoreline slope, low water levels in spring can provide suitable conditions for downslope germination and seedling growth of hygrophytes and emergent species38, while high water levels in summer can promote the spread of aquatic plants, as well as prevent terrestrial plant invasion and lake swamping. Tis approach took into account the lake topographic fea- tures, as well as the aquatic plants’ growth characteristics. One key aspect of this method was to use vegetation coverage to represent the lake ecosystem health. By means of a response relation analysis between vegetation coverage and germination water level, the vegetation coverage and germination zones of aquatic plants could be predicted once the germination water level has been determined. Te advantage of this methodology for estimating the EWLs is that it is cost-efective and utilises existing data that are available for most shallow lakes. Te underwater terrain data of the lake is required in this method, and it can be obtained by feld monitoring and remote sensing images39–41. It should be noted that the recommendations for EWLs are made under the assumption that vegetation coverage is a good indicator for evaluating the health of lake ecosystems. Tis assumption relies heavily on the prior knowledge about the water level requirements of aquatic plants. However, due to the complexity of ecosystems, the quantitative relationship between water levels and aquatic plants is not yet completely clear42,43. Further knowledge about the water level requirements of aquatic plants is still required to enable more accurate predictions regarding the vegetation coverage in shallow lakes. Although this article only mentioned aquatic plants, the method can also be extended to other aquatic species, and it depends on the ecological restoration goals of the lake ecosystem. Te recommended EWLs can be deter- mined if the water level requirements of aquatic species are available. Te methodology developed in this paper is simple and can be conducted over a short time frame. It can provide a reliable reference for water level regulation of shallow lakes especially the lakes with insufcient data.

Received: 19 June 2019; Accepted: 21 January 2020; Published: xx xx xxxx

References 1. Silow, E. et al., Recent efects of human activities on lake Baikal ecosystem. (2015). 2. Ye, X. et al. Quantifying the Human Induced Water Level Decline of China’s Largest Freshwater Lake from the Changing Underlying Surface in the Lake Region. Water Resources Management. 32, 1467–1482 (2018). 3. Jia, B. Y. et al. Driving Efect of Human Activity on the Environmental Change of the Sancha Lake. in International Conference on Biomedical Engineering & Biotechnology. (2012). 4. Cooke, G. D., Welch, E. B. & Peterson, S. A. Lake and Reservoir Restoration. Freshwater Science. (1986). 5. Coops, H. & Hosper, S. H. Water-level Management as a Tool for the Restoration of Shallow Lakes in the Netherlands. Lake and Reservoir Management. 18, 293–298 (2002). 6. Casanova, M. T. & Brock, M. A. How do depth, duration and frequency of fooding infuence the establishment of wetland plant communities? Plant Ecology. 147, 237–250 (2000). 7. Bunn, S. E. et al., National Wetlands R&D Program: Scoping Review. Land and Water Resources Research and Development Corporation (1997). 8. Hassanzadeh, E., Zarghami, M. & Hassanzadeh, Y. Determining the Main Factors in Declining the Urmia Lake Level by Using System Dynamics Modeling. Water Resources Management. 26, 129–145 (2012). 9. Dai, L. et al. Optimal operation of the Tree Gorges Reservoir subject to the ecological water level of . Environmental Earth Sciences. 75, 1–14 (2016).

Scientific Reports | (2020)10:5637 | https://doi.org/10.1038/s41598-020-62454-5 8 www.nature.com/scientificreports/ www.nature.com/scientificreports

10. Guan, L. et al. Optimizing the timing of water level recession for conservation of wintering geese in Dongting Lake, China. Ecological Engineering. 88, 90–98 (2016). 11. Leira, M. & Cantonati, M. Efects of water-level fuctuations on lakes: an annotated bibliography. Hydrobiologia. 613, 171–184 (2008). 12. Wantzen, K. M. et al. Ecological efects of water-level fuctuations in lakes: an urgent issue. Hydrobiologia. 613, 1–4 (2008). 13. Shekar, M. et al. Flooding events and rising water temperatures increase the signifcance of the reed pathogen Pythium phragmitis as a contributing factor in the decline of Phragmites australis. Hydrobiologia. 613, 109–115 (2008). 14. Valk, A. G. V. D., Squires, L. & Welling, C. H. Assessing the Impacts of an Increase in Water Level on Wetland Vegetation. Ecological Applications. 4, 525–534 (1994). 15. Magee, T. K. & Kentula, M. E. Response of wetland plant species to hydrologic conditions. Wetlands Ecology & Management. 13, 163–181 (2005). 16. Zhang, X., Liu, X. & Wang, H. Developing water level regulation strategies for macrophytes restoration of a large river-disconnected lake, China. Ecological Engineering. 68, 25–31 (2014). 17. Arthington, A. H. & Zalucki, J. M. Comparative Evaluation of Environmental Flow Assessment Techniques: Review of Methods. Land and Water Resources Research and Development Corporation. (1998). 18. Chen, Y. et al. Determination of the ecological water-level and assuring degree in the Lake Gaoyou,northern Jiangsu with long-term hydrological alteration. Journal of Lake Sciences. 29, 398–408 (2017). 19. Gao, J. et al. Minimum ecological water depth of a typical stream in Taihu Lake Basin, China. Quaternary International. 226, 136–142 (2010). 20. Shuler, S. W. & Nehring, R. B. Using the Physical Habitat Simulation Model to Evaluate a Stream Habitat Enhancement Project. Rivers. 4, 175–193 (1994). 21. Xu, Z. et al. Minimum Ecological Water Requirements for Lakes Taking in - Sending Out. Water. Resources Science. 27, 140–144 (2005). 22. Gore, J. A. & Nestler, J. M. Instream fow studies in perspective. Regulated Rivers Research & Management. 2, 93–101 (1988). 23. Burgess, G. K. & Vanderbyl, T. L. Habitat Analysis Method for Determining Environmental Flow Requirements. In Institution of Engineers. Australia (1996). 24. Cui, B. et al. Estimation of ecological water requirements based on habitat response to water level in Huanghe River Delta, China. Chinese Geographical Science. 20, 318–329 (2010). 25. Liu, X. et al. A novel methodology for the assessment of water level requirements in shallow lakes. Ecological Engineering. 102, 31–38 (2017). 26. Gan, F. et al. New method and application of estimating ecological water level of the Lake Poyang. Journal of Lake Sciences. 27, 783–790 (2015). 27. Xu, Z., Chen, M. & Dong, Z. Researches on the calculation methods of the lowest ecological water level of lake. Acta Ecologica Sinica. 24, 2324–2328 (2004). 28. Yang, Z., et al., Te theory, method and practice of the ecological water demand. Beijing: Science Press. Beijing: Science Press (2003). 29. Yan, W., Liu, L. & Wang, C. Simplifed Solution for Minimum Ecological Water Demand in Freshwater Lake and Its Application. Water Resources & Power. 30, 6–8 (2012). 30. Cheng, J., et al. Study on the Lowest Ecological Water Level of East Dongting Lake. Jiangxi Science. 932–937 (2015). 31. Jia, Q., et al. Study on biomass dynamics of Phragmites communis community in Panjin wetland. Journal of Meteorology & Environment. 53–58 (2006). 32. Qiao, B. et al. Study on Growth Characteristics of Reed and Its Habitat Soil Factors in Typical Lake-Wetland of Yinchuan. Acta Botanica Boreali-Occidentalia Sinica. 37, 569–577 (2017). 33. Spence, D. H. N. Te Zonation of Plants in Freshwater Lakes. Advances in Ecological Research. 12, 37–125 (1982). 34. Arthington, A. H. & Pusey, B. J. In-stream fow management in Australia: methods, defciencies and future directions. Australian Biologist. 6, 52–60 (1993). 35. Zhang, Y. et al. Aquatic vegetation in response to increased eutrophication and degraded light climate in Eastern Lake Taihu: Implications for lake ecological restoration. Scientifc Reports. 6, 1–12 (2016). 36. Wu, X. Efects of water level and harvesting on submerged macrophytes growth. Nanjing Normal University (2012). 37. Liu, Y. et al. Role of water level fuctuation on aquatic vegetation in lakes. 26, 3117–3126 (2006). 38. Coops, H. & Velde, G. V. D. Seed dispersal, germination and seedling growth of six helophyte species in relation to water-level zonation. Freshwater Biology. 34, 13–20 (1995). 39. Lyzenga, D. R. Passive remote sensing techniques for mapping water depth and bottom features. Applied Optics. 17, 379 (1978). 40. Jagalingam, P., Akshaya, B. J. & Hegde, A. V. Bathymetry Mapping Using Landsat 8 Satellite Imagery. Procedia Engineering. 116, 560–566 (2015). 41. Miecznik, G. & Bader, B. W. Bathymetric techniques using satellite imagery. (2017). 42. Wei, H., Cheng, S. & Wu, Z. Efects of Hydrological Characteristics on Aquatic Plants. Modern Agricultural Sciences & Technology. 13–16 (2010). 43. Tong, G. & He, C. Assessing the ecological role of water level fuctuations on sedimentary information in a shallow lake. Fundamental & Applied Limnology. 191, 223–237 (2018).

Acknowledgements We thank the National Key Research and Development Program (No. 2017YFA0603704 and No. 2017YFC1502500). We are grateful for the feedback from the handling editor and anonymous reviewers. Author contributions W.Y., M.X.X. and L.P.Z. conceived the idea and collected the data for the study. W.Y., M.X.X. and R.Q.L. conducted feldwork and performed the analyses. W.Y. and Q.L.D. wrote the manuscript, with constructive feedback and additions from all authors. Competing interests Te authors declare no competing interests. Additional information Correspondence and requests for materials should be addressed to W.Y. or L.Z. Reprints and permissions information is available at www.nature.com/reprints.

Scientific Reports | (2020)10:5637 | https://doi.org/10.1038/s41598-020-62454-5 9 www.nature.com/scientificreports/ www.nature.com/scientificreports

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional afliations. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Cre- ative Commons license, and indicate if changes were made. Te images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not per- mitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.

© Te Author(s) 2020

Scientific Reports | (2020)10:5637 | https://doi.org/10.1038/s41598-020-62454-5 10